Portal de Periódicos da CAPES

High-frequency oscillatory ventilation and an interventional lung assist device to treat hypoxaemia and hypercapnia

2004; Elsevier BV; Volume: 93; Issue: 4 Linguagem: Inglês

10.1093/bja/aeh231

1471-6771

Matthias David, Wolfgang Heinrichs,

Cardiac Arrest and Resuscitation

A male patient accidentally aspirated paraffin oil when performing as a fire-eater. Severe acute respiratory distress syndrome ( Pa2/F 2ratio 10.7 kPa) developed within 24 h. Conventional pressure-controlled ventilation (PCV) with high airway pressures and low tidal volumes failed to improve oxygenation. Hypercapnia ( Pa212 kPa) with severe acidosis (pH<7.20) ensued. Treatment with high-frequency oscillatory ventilation (HFOV) and a higher adjusted airway pressure (35 cm H2O) improved the Pa2/F 2ratio within 1 h from 10.7 to 22.9 kPa, but the hypercapnia and acidosis continued. Stepwise reduction of the mean airway pressure (26 cm H2O), and oscillating frequencies (3.5 Hz), as well as increasing the oscillating amplitudes (95 cm H2O) resulted in an unchanged Pa2, but oxygenation worsened. The new pumpless extracorporeal interventional lung assist device (ILA, NovaLungR, Hechingen, Germany) was therefore used for carbon dioxide elimination to enable a less aggressive ventilation strategy. Pa2normalized after initiation of ILA. HFOV with a mean airway pressure of 32 cm H2O was maintained, but with a higher oscillatory frequency (9 Hz) and very low oscillatory amplitude (25 cm H2O). After 6 days, the patient was transferred to a conventional ventilator, and ILA was discontinued after 13 days without complications. A male patient accidentally aspirated paraffin oil when performing as a fire-eater. Severe acute respiratory distress syndrome ( Pa2/F 2ratio 10.7 kPa) developed within 24 h. Conventional pressure-controlled ventilation (PCV) with high airway pressures and low tidal volumes failed to improve oxygenation. Hypercapnia ( Pa212 kPa) with severe acidosis (pH<7.20) ensued. Treatment with high-frequency oscillatory ventilation (HFOV) and a higher adjusted airway pressure (35 cm H2O) improved the Pa2/F 2ratio within 1 h from 10.7 to 22.9 kPa, but the hypercapnia and acidosis continued. Stepwise reduction of the mean airway pressure (26 cm H2O), and oscillating frequencies (3.5 Hz), as well as increasing the oscillating amplitudes (95 cm H2O) resulted in an unchanged Pa2, but oxygenation worsened. The new pumpless extracorporeal interventional lung assist device (ILA, NovaLungR, Hechingen, Germany) was therefore used for carbon dioxide elimination to enable a less aggressive ventilation strategy. Pa2normalized after initiation of ILA. HFOV with a mean airway pressure of 32 cm H2O was maintained, but with a higher oscillatory frequency (9 Hz) and very low oscillatory amplitude (25 cm H2O). After 6 days, the patient was transferred to a conventional ventilator, and ILA was discontinued after 13 days without complications. The ARDS Net trial has shown that the ventilation strategy used to manage patients with acute respiratory distress syndrome (ARDS) affects outcome.1Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome The Acute Respiratory Distress Syndrome Network.N Engl J Med. 2000; 342: 1301-1308Crossref PubMed Scopus (10087) Google Scholar High inspiratory pressures with large tidal volumes, leading to over-distension of the lung and cyclic alveolar collapse with reopening during inspiration, have been identified as potential triggers of ventilator-induced lung injury.2Dreyfuss D Soler P Basset G Saumon G. High inflation pressure pulmonary edema. Respective effects of high airway pressure, high tidal volume, and positive end-expiratory pressure.Am Rev Respir Dis. 1988; 137: 1159-1164Crossref PubMed Scopus (1320) Google Scholar The current strategy for conventional mechanical ventilation in ARDS is to prevent further lung injury. High-frequency oscillatory ventilation (HFOV) allows the application of higher mean airway pressures with very low tidal volumes, compared with conventional pressure-controlled ventilation (PCV) modes. HFOV is a safe and effective mode of respiratory support in the treatment of adult patients with ARDS.3Derdak S Mehta S Stewart TE Multicenter Oscillatory Ventilation For Acute Respiratory Distress Syndrome Trial (MOAT) Study Investigators High-frequency oscillatory ventilation for acute respiratory distress syndrome in adults.Am J Respir Crit Care Med. 2002; 166: 801-808Crossref PubMed Scopus (514) Google Scholar, 4David M Weiler N Heinrichs W et al.High-frequency oscillatory ventilation in adult acute respiratory distress syndrome.Intensive Care Med. 2003; 29: 1656-1665Crossref PubMed Scopus (85) Google Scholar However, the use of high airway pressures in combination with an inspiratory pressure limitation may lead to insufficient carbon dioxide removal, resulting in severe respiratory acidosis. A new supportive therapy for use in ARDS patients is the Interventional Lung Assist device (ILA, NovaLung®, Hechingen, Germany), which removes carbon dioxide from the blood. This system contains a specially designed low resistance lung membrane, which uses the pressure difference between the arterial and venous circulation as the driving force for blood flow. The extracorporeal blood flow is approximately 25% of the cardiac output. This system enables the use of high airway pressures for oxygenation in combination with very low tidal volumes to avoid ventilator-induced lung injury, and this gains time for lung recovery. A male patient (age 30 yr, height 170 cm, weight 57 kg, idealized body weight (IBW) 72 kg, APACHE II score 25) with severe ARDS as a result of aspiration of paraffin oil was treated with a combination of HFOV and a pumpless ILA. The patient accidentally aspirated paraffin oil (40 ml) during a performance as a fire-eater. Immediately after aspiration, he experienced slight dyspnoea and chest pain during inspiration. By 8 h, the patient had severe dyspnoea and the chest X-ray demonstrated pulmonary infiltration in both lower lobes and the right middle lobe. On examination, chest auscultation revealed bilateral dry rales over the lower chest, core body temperature was 39.3°C, invasive arterial pressure was 93/30 mm Hg, heart rate 136 min−1 in sinus rhythm, and central venous pressure 16 mm Hg. Computed tomography of the lung at that time showed bilateral lower lobe consolidation, atelectasis, and infiltration in the right middle lobe. The Pa2was 6.9 kPa. The trachea was intubated and the lungs were ventilated mechanically using a PCV mode. Bronchoscopy revealed solitary blood patches in the central bronchial system and brownish secretion in both lower lobes. Diagnostic bronchoalveolar lavage (40 ml of normal saline) produced dark debris, which smelled of paraffin oil. Despite increased positive end-expiratory pressure (PEEP; 20 cm H2O), mean airway pressure (Pmean; 27 cm H2O), and repeated recruiting manoeuvres (continuous positive airway pressure level of 45 cm H2O for at least 30 s), oxygenation did not improve over 2 h. The Pa2(kPa)/ F 2ratio was 10.7 kPa, oxygenation index (Pmean× F 2×100/ Pa2(mm Hg)) 32, and hypercapnia ( Pa212 kPa, arterial pH 7.20, arterial base excess −4.0 mmol litre−1) was evident. HFOV (SensorMedics 3100B, Yorba Linda, CA, USA) was started, which included an initial lung recruitment strategy. The first setting was an increase of the mean airway pressure of 5 cm H2O compared with the last measured mean airway pressure during PCV, an oscillatory frequency of 5 Hz, an inspiratory time of 33% of the respiratory cycle, a flow bias of 30 litre min−1, and a F 2of 1.0. The arterial oxygen tension and the arterial carbon dioxide tension were monitored continuously with an in-line arterial multiparameter sensor (Partrend 7, Diametrics Medical Ltd, England). The oscillating pressure amplitude was scaled to the online measurement of Pa2. With HFOV, lung volume was then recruited by a stepwise increase (2 cm H2O) of the adjusted mean airway pressure up to a maximum of 40 cm H2O. After each step, Pa2and its trend were analyzed, and mean airway pressure was increased as long as Pa2increased. If Pa2decreased when increasing the mean airway pressure, we also analyzed the trend in Pa2. Any decrease in Pa2without an initial positive trend was interpreted as no further lung recruitment and over-distension of open lung units. Mean airway pressure was then reduced. After achieving maximum recruitment using this strategy, the lowest possible mean airway pressure was selected that would keep the lung open. This was determined by stepwise reductions in mean airway pressure to the point where collapse of alveolar units became evident from a decrease in Pa2. Mean airway pressure was then set at 2–3 cm H2O above this pressure. Oxygenation improved during HFOV within 1 h ( Pa2/F 2ratio 22.9 kPa; oxygenation index 17) and remained stable thereafter (Fig. 1). However, Pa2increased again to 13.3 kPa within 24 h, with concomitant respiratory acidosis (arterial pH 7.14, arterial base excess −4 mmol litre−1). Neither lower oscillatory frequencies of 3.5 Hz, nor maximum oscillatory amplitudes lowered the Pa2levels. Stepwise reduction of mean airway pressure (2 cm H2O every 30 min) from 32 to 26 cm H2O led to only a slight decrease of Pa2, but to a major decrease in oxygenation ( Pa2/F 2ratio from 26 to 12 kPa). We decided to establish the ILA for carbon dioxide removal to offer less aggressive ventilation with HFOV (high mean airway pressures with lower oscillatory pressure amplitudes and high oscillatory frequencies). We used the left femoral artery and the right femoral vein (Seldinger technique) for insertion of a 17 French gauge arterial and 19 French gauge venous cannula. After insertion, the pre-filled (isotonic saline) ILA was connected; initial passive blood flow was 2.1 litre min−1, and gas flow (oxygen) in the membrane lung was 12 litre min−1. Because of the heparin bonded system, systemic anticoagulation with heparin was targeted only to an activated clotting time of 120–140 s. The patient tested negative for heparin-induced thrombocytopenia (HIT II). After an additional recruitment, HFOV was maintained with a mean airway pressure of 32 cm H2O, whereas oscillatory frequency was increased (from 3.5 to 9 Hz). Oscillatory amplitude was decreased from 95 to 25 cm H2O to reduce tidal volumes. Immediately after ILA initiation, Pa2fell from 13.3 (arterial pH 7.14, arterial base excess −5 mmol litre−1) to 9.3 kPa (arterial pH 7.22, arterial base excess −2 mmol litre−1) and returned to normal within 4 h. The Pa2/F 2ratio was unchanged after ILA started and remained higher than 20 kPa (Fig. 1) with the HFOV setting used. Blood flow through the extracorporeal system was measured by ultrasound (Blood Flow Monitoring System, NovaLung®, Hechingen, Germany). Mean arteriovenous shunting with the ILA during treatment was 24 (4)% of the cardiac output, measured by a thermodilution technique. The targeted mean arterial pressure was 70 mm Hg and the minimal cardiac index to ensure sufficient blood flow through the membrane lung was 2.5 litre min−1 m−2. Norepinephrine was only used for 13 h (maximum dosage: 1.5 μg kg−1 min−1) after starting ILA to reach this index. Figure 2 shows the measured blood flow and adjusted gas flow (oxygen) during ILA treatment. We were able first to reduce F 2stepwise, and then mean airway pressure on HFOV over several days. The patient was transferred to a conventional ventilator on day 7 with PCV, and on day 9 switched to a pressure support ventilation mode. Tidal volume was targeted to less than 7 ml kg−1 IBW (25×(height in metres)2Dreyfuss D Soler P Basset G Saumon G. High inflation pressure pulmonary edema. Respective effects of high airway pressure, high tidal volume, and positive end-expiratory pressure.Am Rev Respir Dis. 1988; 137: 1159-1164Crossref PubMed Scopus (1320) Google Scholar). The patient improved and the performance of the ILA was adjusted by reducing the gas flow (oxygen) of the membrane lung. The ILA was stopped 13 days after initiation after a successful trial without gas flow (oxygen) from the membrane lung. We observed no ILA related adverse events during treatment despite reduced systemic anticoagulation for percutaneous tracheostomy on day 12. Weaning from ventilation was interrupted by nosocomial pneumonia (Staphylococcus aureus and Stenotrophomonas maltophilia) on day 20, which resolved. After 43 days, the tracheostomy was removed. The patient suffered from acute renal failure on admission and required renal replacement therapy for 30 days. The patient was discharged from hospital on day 52 for rehabilitation. This patient, with severe ARDS after aspiration of paraffin oil, is the first description of the simultaneous use of two unconventional therapies forms: HFOV and the new ILA. Morris and colleagues failed to demonstrate, in a randomized clinical trial, any benefit using pump driven venovenous extracorporeal membrane oxygenation (vvECMO), compared with conventional therapy in ARDS patients.5Morris AH Wallace CJ Menlove RL et al.Randomized clinical trial of pressure controlled inversed ratio ventilation and extracorporeal CO2 removal for adult respiratory distress syndrome.Am J Respir Crit Care Med. 1994; 149: 295-305Crossref PubMed Scopus (754) Google Scholar They used high PEEP levels in both groups (vvECMO and conventional therapy), and very high peak airway pressures (45.4 (1.7) cm H2O) in the ECMO group. A serious problem was the high incidence of bleeding and requirement for blood products, as a result of the high-dose heparin used during vvECMO. Only animal studies of pumpless extracorporeal carbon dioxide removal have been published in the last two decades, together with more recent case reports in a few patients with ARDS. The latter focus on treatment with the new pumpless ILA system.6Gattinoni L Kolobow T Agostoni A et al.Clinical application of low frequency pressure controlled ventilation with extracorporeal CO2 (LFPPV-ECCO2-R) removal in treatment of adult respiratory distress syndrome (ARDS).J Artif Organs. 1979; 2: 282-283Google Scholar7Awad JA Cloutier R Fournier L et al.Prolonged pumpless arteriovenous perfusion for carbon dioxide extraction.Ann Thorac Surg. 1995; 51: 530-540Google Scholar8Cappelletti DD Olshove V Tallmann Jr., RD Pumpless arterial-venous extracorporeal CO2 removal during acute lung injury.J Extra-Corpor Technol. 1996; 28: 6-12Google Scholar9Zwischenberger JB Alpard SK Conrad SA Johnigan RH Bidani A. Arteriovenous carbon dioxide removal: development and impact on ventilator management and survival during severe respiratory failure.Perfusion. 1999; 14: 299-310Crossref PubMed Scopus (13) Google Scholar10Reng M Philipp A Kaiser M Pfeiffer M Gruene S Schoelmerich J. Pumpless extracorporeal lung assist and adult respiratory distress syndrome.Lancet. 2000; 15: 219-220Abstract Full Text Full Text PDF Scopus (100) Google Scholar, 11Liebold A Reng CM Philipp A Pfeiffer M Birnbaum DE. Pumpless extracorporeal lung assist—experience with the first 20 cases.Eur J Cardiothorac Surg. 2000; 17: 608-613Crossref PubMed Scopus (75) Google Scholar Typical complications associated with the ILA device were oxygenator failure, cannula problems and thrombus formation. However, Liebold reported 20 patients with ARDS, where conventional mechanical ventilation had failed and pumpless extracorporeal lung assist was established. They found no serious adverse events (such as bleeding, ischaemia of the punctured legs, thromboembolism) related to the extracorporeal therapy, and the system was effective for carbon dioxide removal.11Liebold A Reng CM Philipp A Pfeiffer M Birnbaum DE. Pumpless extracorporeal lung assist—experience with the first 20 cases.Eur J Cardiothorac Surg. 2000; 17: 608-613Crossref PubMed Scopus (75) Google Scholar We recorded no further improvement in oxygenation after ILA started, but carbon dioxide removal was effective. In respect of oxygenation, blood flow through the membrane lung was the limiting factor, reaching only 28% of the cardiac output and at no time being equal to pump driven systems. The effect of HFOV on oxygenation and its safety in patients with ARDS has been documented in several studies.3Derdak S Mehta S Stewart TE Multicenter Oscillatory Ventilation For Acute Respiratory Distress Syndrome Trial (MOAT) Study Investigators High-frequency oscillatory ventilation for acute respiratory distress syndrome in adults.Am J Respir Crit Care Med. 2002; 166: 801-808Crossref PubMed Scopus (514) Google Scholar, 4David M Weiler N Heinrichs W et al.High-frequency oscillatory ventilation in adult acute respiratory distress syndrome.Intensive Care Med. 2003; 29: 1656-1665Crossref PubMed Scopus (85) Google Scholar The HFOV allowed application of a constant high mean airway pressure, and avoided high cyclic inspiratory pressures as well as end-expiratory lung collapse with reopening on inspiration. If hypercapnia is evident with HFOV, reduced oscillatory frequencies, increased oscillatory amplitudes, and increased bias flow are used to decrease it. However, permanent administration of oscillatory frequencies below 4 Hz in combination with high oscillating amplitudes results in higher tidal volumes, which cancels the benefit of the HFOV treatment. So called ‘permissive hypercapnia’ is one element of accepted protective lung ventilatory strategies, but only if the target primarily is to avoid end-expiratory lung collapse and to reduce alveolar stretch with low tidal volumes. No clear evidence demonstrates that moderate respiratory acidosis worsens the condition of critical ill patients, and the safe level of hypercapnia and respiratory acidosis remains unknown. We used the new pumpless ILA for carbon dioxide removal in a clinical situation where oxygenation improved only with high mean airway pressures, whereas adjusted ventilator settings were insufficient to control the resulting hypercapnic acidosis. The ILA was only supportive therapy to enable less aggressive ventilator settings and not a ventilator replacement. The application of high airway pressures to maintain oxygenation during ILA is possible with either conventional PCV modes or with HFOV. We continued HFOV during ILA therapy because of the improvement in the Pa2/F 2ratio after starting it, but used high oscillatory frequencies and low oscillatory amplitudes to achieve very low tidal volumes.